Build a SaaS churn prediction tool with these features:

1. Data integration: Connect to your SaaS app's database to pull user activity logs, login frequency, feature usage, session duration, and billing history. Support CSV upload for initial setup.

2. Churn risk scoring: AI analyzes usage patterns, engagement decline, support ticket frequency, and billing data to assign each account a churn risk score (Low, Medium, High, Critical). Update scores daily.

3. Risk factor breakdown: For each at-risk account, show the top contributing factors (e.g. login frequency down 60%, no core feature usage in 14 days, support ticket unresolved).

4. AI retention recommendations: For each flagged account, generate a specific retention strategy (e.g. 'Send a re-engagement email highlighting unused feature X', 'Offer a discount on next billing cycle', 'Schedule a check-in call').

5. Alert system: Notify account managers via email or Slack when an account moves to High or Critical risk. Include the risk factors and recommended action.

6. Cohort analysis dashboard: Visualize churn risk across user cohorts (signup month, plan tier, industry, team size). Spot patterns in which segments churn most.

7. Retention tracking: Log retention actions taken and track whether the account was saved or lost. Build a feedback loop to improve predictions over time.

8. Health score timeline: Show each account's health score over time so you can see the decline leading up to a churn event.

9. Custom risk thresholds: Let admins configure what counts as High vs Critical risk based on their business context.

10. Weekly churn report: Automated weekly email summarizing new at-risk accounts, actions taken, and accounts saved.

Use a data-dense but clean dashboard UI. Prioritize the at-risk account list with clear risk indicators and recommended actions.